Diagnostic & prognostic role of microRNAs in paediatric acute myeloid leukaemia

Introduction

Paediatric acute myeloid leukaemia (AML) is a heterogeneous disease in terms of diverse cytogenetic and molecular abnormalities, all leading to malignant transformation of haematopoietic progenitors12. It accounts for almost 15-20 per cent of all paediatric leukaemia. Although there has been a significant improvement in the overall survival rate of paediatric AML patients from 30 to 73 per cent, still nearly half of them relapse3. Therefore, diagnostic and prognostic biomarkers for classifying different risk groups as well as more effective molecular targeted therapies are urgently needed for better management of paediatric AML patients.

MicroRNAs (miRNAs) are a group of non-coding RNAs, which mainly function through complementary base pairing to the 3’ untranslated region of target messenger RNA (mRNA), followed by degradation of mRNA and/or translational inhibition4. These miRNAs, now recognized as epigenetic biomarkers, play vital functions in numerous cellular events ranging from organogenesis to immunity56. Deregulation of miRNAs affects normal cell growth and development including multiple cellular events including cell cycle regulation, differentiation and cell death, leading to various diseases including cancer7. The expression pattern of miRNA has been studied in adult AML patients, where abnormal expression of different miRNAs has been detected in distinct adult AML subtypes leading to activation or inhibition of essential pathways in leukaemogenesis8. Though information on diagnostic, prognostic and functional importance of miRNAs expression is available in adult AML8, in paediatric AML, information about miRNA expression has been gathered only in a limited series of patients so far9101112131415161718. Hence, in the present review, efforts have been made to summarize the findings of studies published so far in the area of miRNAs in paediatric AML.

MicroRNA expression profile in paediatric acute myeloid leukaemia: A diagnostic and prognostic tool

Table I summarizes studies which have explored the efficacy of miRNA expression as a diagnostic and prognostic biomarker in paediatric AML patients9101112131415161718. Most of these studies have tried to explore the possibilities of identifying specific miRNAs expression or miRNAs expression signature from bone marrow which can distinguish paediatric AML from normal controls. Except two studies in which miRNA microarray platform was used for profiling of human mature miRNAs914, in all other studies quantitative reverse-transcriptase polymerase chain reaction (qRT-PCR) using miRNA-specific stem-loop primers and probes was used for the analysis of miRNA expression (Table I). The miRNA expression was detected using bone marrow samples derived from paediatric AML patients ranging from a minimum of 68 patients13 to a maximum of 169 patients10. The numbers of normal controls were comparatively less, which could be due to ethical considerations involved in bone marrow aspiration from this group of subjects. Despite having variations in methodology and patients recruited and miRNA assayed amongst all the studies, a common feature was dysregulation in miRNA expression in paediatric AML patients when compared with normal controls. Specific miRNAs such as miR-125b910, miR-100911, miR-99a913, miR-37515 and miR-335917 were highly expressed, while miR-29a16 and miR-66318 were underexpressed in paediatric AML patients when compared to normal controls. Using miRNA microarray platform, Zhang et al9 found upregulation of 17 miRNAs and downregulation of 18 miRNAs in paediatric AML patients as compared to normal controls. The expression of a few of these miRNAs was further validated by qRT-PCR, which confirmed the upregulation of miR-99a, miR-100, miR-125b, miR-146a and miR-335 in paediatric AML patients9. Table II lists those miRNAs, which were differentially expressed in the bone marrow of paediatric AML patients as compared to controls.

Table I A summary of studies on miRNAs expression profiling in paediatric acute myeloid leukaemia (AML)

Table II List of microRNAs (miRNAs) found to be dysregulated in paediatric acute myeloid leukaemia (AML)

The impact of aberrant miRNA expression has been studied on clinical and therapeutic outcome in adult AML8. In line with this, a few studies also explored the utility of miRNA expression as a potential prognostic indicator in paediatric AML910111213151617 (Table II). These studies tried to establish a correlation between miRNA expression pattern and clinical outcome in terms of therapeutic response, overall survival and relapse in paediatric AML patients. Zhang et al9 concluded that miR-125b and miR-126a can predict favourable prognosis for M3 and M2 AML patients, respectively. However, the study failed to establish any correlation between miRNA expression and central nervous system (CNS) relapse. In another study by Zhang et al10, expression of miR-125b decreased in paediatric AML patients who achieved complete remission post-therapy, but the levels remained high in patients with relapse. Similar results were also observed in another study by Zhang et al13 using miR-99a. The increased expression of miR-100 has been shown to be associated with poor relapse-free and overall survival as well as unfavourable day 7 response to induction chemotherapy in paediatric AML patients11. However, another group failed to establish any correlation between miR-196a/b expression and overall survival12. Further, increased expression of miR-37515 and miR-33517 and low expression of miR-29a16 in paediatric AML patients have been shown to be associated with poor relapse-free survival and short overall survival in both univariate and multivariate analyses.

The above-mentioned data suggest that quantification of miRNA expression may have clinical utility for risk assessment in paediatric AML. However, as the literature on the expression of miRNAs in paediatric AML is very limited, it is not feasible to evaluate their true potential as an effective diagnostic and prognostic biomarker. What is clear from these studies is that miRNA signatures or expression of specific miRNA is not consistent among different studies. This lack of homogeneity in the findings could be attributed to a number of factors, such as (i) variability in the recruitment of number of patients and controls, (ii) variability in the frequency of distribution of paediatric AML patients in various cytogenetic or genetic groups, (iii) use of unselected mononuclear cells from the bone marrow of healthy donors instead of purified CD34⁺ cells, and (iv) different methods for profiling of miRNA.

Correlation of miRNA expression with cytogenetics in paediatric acute myeloid leukaemia

The heterogeneity and complexity of AML in terms of various cytogenetic abnormalities is well established2. Although many of these abnormalities are rare, some of these occur frequently and are clinically very important as these have been found to be associated with therapeutic outcome and survival2. The commonly detected cytogenetic abnormalities have been classified according to the prognostic information they carry. For example, t(8;21)(q22;q22) and inv(16)(p13.1q22) or t(16;16)(p13.1;q22) and t(15;17)(q22;q12-q21) confer a relatively favourable outcome. However, patients with balanced translocations involving band 11q23 and the myeloid/lymphoid leukaemia (MLL) gene [t(v;11) (v;q23)/MLL] other than t(9;11)(p22;q23), inv(3)(q21q26.2) or t(3;3)(q21;q26.2), t(6;9)(p23;q34), deletion or loss of 5q, monosomy 7, structural alterations of 17p or a complex karyotype [defined as more than or equal to three chromosome aberrations in the absence of t(8;21), inv(16) or t(16;16), t(15;17), t(9;11), t(v;11), t(6;9) and inv(3) or t(3;3)] have a very poor prognosis28. Patients with other chromosome aberrations such as t(9;11) (p22;q23); those not classified as favourable or unfavourable or those with cytogenetically normal AML (CN-AML) are classified as having an intermediate prognosis2.

Using miRNA expression profiling, only a couple of studies have looked at the correlation between specific miRNA signatures and cytogenetic subtypes of paediatric AML1214. Specific miRNAs, which are either upregulated or downregulated in various cytogenetic abnormalities, are listed in Table III. In one of the studies, miR-196a/b was upregulated, whereas miR-29a was downregulated in paediatric AML patients carrying MLL gene rearrangements12. In the same study, a relatively low expression of miR-196a/b was observed in patients with t(8;21), inv(16) and t(15;17) as compared to all other patients. However, there was no correlation between miR-155 expression and specific cytogenetic abnormalities12. Daschkey et al14 found that miR-27a, miR-126, miR-150 and miR-223 were significantly highly expressed, while miR-21 was significantly underexpressed in t(8;21)-positive paediatric AML samples as compared to t(15;17)-positive samples. Further, miR-100, miR-125b and miR-181a/b were highly expressed in t(15;17)-positive leukaemia, whereas miR-146a was highly expressed in both t(8;21)-positive and t(15;17)-positive leukaemia as compared to other cytogenetic groups14. They further concluded that a miRNA signature consisting of 22 miRNAs can correctly classify almost 87 per cent of patient samples belonging to various cytogenetic risk groups such as t(8;21), t(15;17) and MLL-rearranged AML14. In other studies it was found that high expression of miR-37515 and miR-33517 was more frequent in paediatric AML patients with unfavourable karyotypes than in with favourable or intermediate karyotypes. In contrast, miR-29a was highly expressed in paediatric AML patients with favourable karyotypes than in with intermediate or unfavourable karyotypes16.

Table III List of microRNAs associated with various cytogenetics and molecular alterations in paediatric acute myeloid leukaemia (AML)

It is to be noted that the detection of aforementioned cytogenetic alterations involves more uniform and standardized protocols across various laboratories and hence is an accepted method for diagnosis, prognosis and management of AML. This is in contrast to two of the most common methods for miRNA profiling - microarray and qRT-PCR, which have been used in various studies for evaluating the diagnostic and prognostic potential of miRNA in paediatric AML. In miRNA microarray, there is a hybridization between specific miRNA sequences and their respective complementary probes on a slide, thereby producing fluorescent signals, which is measured as distinct spots. As miRNAs are very small and many of these belong to the same family, thereby differing only by a few nucleotides, designing probes for specific miRNAs is a very difficult process and can influence the end result. Differences in the methods of probe designing, probe labelling, hybridization and use of different microarray platforms can also influence the relative abundance of miRNA. Further, qRT-PCR is also a highly variable method and factors such as (i) use of different endogenous miRNA control for normalization, (ii) use of different chemistries such as SYBR-Green, TaqMan probe and molecular beacons, and (iii) sensitivity of different platforms of real-time PCR machines can influence the end-point measurement of expression of specific miRNAs. Thus, it becomes crucial to optimize various methods for the measurement of miRNA expression before integrating them in routine clinical settings for diagnosis, prognosis or management of paediatric AML.

Correlation of miRNA expression with molecular markers in paediatric acute myeloid leukaemia

Nearly half of the AML patients do not have any cytogenetic abnormalities and are classified as CN-AML. However, in various studies it has been found that CN-AML is actually a heterogeneous group with a number of genetic abnormalities, leading to defects in gene expression2. Many of these molecular alterations also carry prognostic information, which makes their analysis clinically very important and viable1920. Molecular alterations such as an internal tandem duplication of the Fms-like tyrosine kinase 3 gene (FLT3-ITD), partial tandem duplication of the MLL gene, mutations of the Wilms tumour 1 (WT1) and high expression of the brain and acute leukaemia, cytoplasmic, erythroblast transformation-specific-related gene and meningioma (MN) (disrupted in balanced translocation) 1 (MN1) genes confer adverse prognosis, whereas mutations in the nucleophosmin (NPM1) and CCAAT/enhancer-binding protein alpha (CEBPA) genes confer favourable prognosis28. Further, numerous combinations of these markers have also provided useful information for predicting clinical outcome of CN-AML patients. In one such example, AML patients positive for NPM1 mutations, but negative for FLT3-ITD, showed a better outcome than patients who were positive for FLT3-ITD, regardless of NPM1 mutations, or have wild-type FLT3 and NPM1 alleles2.

As many of these molecular alterations carry important prognostic information in paediatric AML, it is imperative to establish their correlation with the expression of specific miRNAs (Table III). MiR-196a/b was found to be highly expressed in paediatric AML patients positive for NPM1 mutation or FLT3-ITD, whereas patients carrying CEBPA mutations were having low expression of miR-196a/b12. The authors also observed high miR-155 expression in FLT3-ITD and NPM1-mutated cases12. No correlation could be established between miR-29a expression and FLT3-ITD or neuroblastoma rat sarcoma viral oncogene homolog (N-RAS) and V-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog (K-RAS) mutations, miR-155 expression and N/K-RAS mutations and miR-196a/b expression and presence of WT1 or N/K-RAS mutations12. Further studies are needed to establish a significant association between miRNA expression signatures and specific molecular alteration in paediatric AML.

Conclusions and future directions

The development of sophisticated, high-throughput technologies for profiling miRNAs, such as miRNA microarray platforms, array/card-based qRT-PCR and next-generation sequencing methods for small RNA sequencing, have made it possible not only to study the dysregulated miRNA expression in AML but also to attain a global, multidimensional view of gene regulation by combining it with various DNA, RNA or protein-based high-throughput approaches. However, before being universally accepted, the role of miRNAs in paediatric AML needs to be investigated further and should also be validated in different laboratories working on different cohorts across the world. In addition, use of an adequate number of patients and appropriately matched controls to get significant results, standardization of methods for the collection, storage and processing of biological samples, uniformity in assay platforms and interpretation of huge amount of data, are some of the areas which need to be explored. To get specific miRNAs involved in AML, another challenge would be to get reproducible data sets with highly specific and sensitive statistical numbers. Once validated, various in vivo studies, with appropriate preclinical knock-in and knock-out animal models, may deliver the functional role of various miRNAs and their involvement in leukaemia development. After functional characterization, specific miRNAs can further be used for designing miRNA-based therapeutic strategies. Available preliminary data suggest that miRNA-based therapeutic approach can be a viable option for disease management.